ChipSeg: an automatic tool to segment bacteria and mammalian cells cultured in microfluidic devices

ABSTRACTExtracting quantitative measurements from time-lapse images is necessary in external feedback control applications, where segmentation results are used to inform control algorithms. While such image segmentation applications have been previously reported, there is in the literature a lack of open-source and documented code for the community. We describe ChipSeg, a computational tool to segment bacterial and mammalian cells cultured in microfluidic devices and imaged by time-lapse microscopy. The method is based on thresholding and uses the same core functions for both cell types. It allows to segment individual cells in high cell-density microfluidic devices, to quantify fluorescence protein expression over a time-lapse experiment and to track individual cells. ChipSeg enables robust segmentation in external feedback control experiments and can be easily customised for other experimental settings and research aims.

Towards automated control of embryonic stem cell pluripotency

Mouse embryonic stem cells (mESCs) have been shown to exist in three distinct pluripotent states (ground, naïve and primed pluripotent states), depending on culture conditions. External feedback control strategies have been, so far, mainly used to automatically regulate gene expression in bacteria and yeast. Here, we exploit a microfluidics/microscopy platform and segmentation and external feedback control algorithms for the automatic regulation of pluripotency phenotypes in mESCs. We show feasibility of automatically controlling, in living mESCs, levels of an endogenous pluripotency gene, Rex1, through a fluorescent reporter, used as control output, and drugs commonly used to modulate pluripotency (i.e. MEK kinase and Gsk3β inhibitors) as control inputs. Our results will ultimately aid in the derivation of superior protocols for pluripotency maintenance and differentiation of mouse and human stem cells.